Speed up
in [Matlab]
|
From: Oleg Komarov on 13 Jul 2010 06:08 "Cris Luengo" <cris.luengo(a)google.for.my.name.to.contact.me> wrote in message <i1hcf7$p0r$1(a)fred.mathworks.com>... > "Oleg Komarov" <oleg.komarovRemove.this(a)hotmail.it> wrote in message <i1habo$dfn$1(a)fred.mathworks.com>... > > N = 256; x = randn(N,N,N); > > > > % Elapsed time is 0.546098 seconds. > > tic; y = 0.5*( x(:,:,1:N-1) + x(:,:,2:N)); toc > > > > % Elapsed time is 0.301945 seconds. > > tic; y1 = reshape((reshape(x(:,:,1:N-1),1,[]) + reshape(x(:,:,2:N),1,[]))*.5,256,256,[]); toc > > Oleg, > I'm baffled as to why the reshaped arrays compute faster. I tried it, it is faster, I just don't see why. Please enlighten me! > Cheers, > Cris. Unfortunately I can't give you a detailed explanation but the point is that when ndims(x) > 2 I've seen often that the code isn't optimized (which you can't see for sum or '+'). I hope anybody else can answer you with more details... Oleg
From: Sukanta on 13 Jul 2010 08:39 Thanks a lot folks for all the suggestions. I appreciate it. I am also puzzled by the fact that an explicit for-loop reduces the cost. tic; y = 0.5*( x(:,:,1:N-1) + x(:,:,2:N)); toc Elapsed time is 0.294341 seconds. tic; y = zeros(N,N,N-1); for k = 1:N-1; y(:,:,k) = 0.5*(x(:,:,k)+x(:,:,k+1)); end; toc Elapsed time is 0.235350 seconds. tic; y1 = reshape((reshape(x(:,:,1:N-1),1,[]) + reshape(x(:,:,2:N),1,[]))*.5,N,N,[]);toc Elapsed time is 0.210588 seconds. "Oleg Komarov" <oleg.komarovRemove.this(a)hotmail.it> wrote in message <i1hdu4$pvg$1(a)fred.mathworks.com>... > "Cris Luengo" <cris.luengo(a)google.for.my.name.to.contact.me> wrote in message <i1hcf7$p0r$1(a)fred.mathworks.com>... > > "Oleg Komarov" <oleg.komarovRemove.this(a)hotmail.it> wrote in message <i1habo$dfn$1(a)fred.mathworks.com>... > > > N = 256; x = randn(N,N,N); > > > > > > % Elapsed time is 0.546098 seconds. > > > tic; y = 0.5*( x(:,:,1:N-1) + x(:,:,2:N)); toc > > > > > > % Elapsed time is 0.301945 seconds. > > > tic; y1 = reshape((reshape(x(:,:,1:N-1),1,[]) + reshape(x(:,:,2:N),1,[]))*.5,256,256,[]); toc > > > > Oleg, > > I'm baffled as to why the reshaped arrays compute faster. I tried it, it is faster, I just don't see why. Please enlighten me! > > Cheers, > > Cris. > > Unfortunately I can't give you a detailed explanation but the point is that when > ndims(x) > 2 I've seen often that the code isn't optimized (which you can't see for sum or '+'). > I hope anybody else can answer you with more details... > > Oleg
First
|
Prev
|
Pages: 1 2 Prev: OuterPosition in absolute units and plotyy Next: Error Embedded Matlab |